Pkgndep: a tool for analyzing dependency heaviness of R packages

Area of Science:

• Bioinformatics
• Computational Biology
• Software Engineering

Background:

• The proliferation of R packages for bioinformatics has led to complex interdependencies.
• Managing package dependencies is crucial for reproducible and efficient analysis.

Purpose of the Study:

• To introduce a novel metric, dependency heaviness, for quantifying package dependencies.
• To develop tools and perform analyses to identify and mitigate dependency complexity in R ecosystems.

Main Methods:

• Developed the 'dependency heaviness' metric to measure unique parent package contributions.
• Implemented the metric in the R package 'pkgndep'.
• Conducted a global analysis of dependency heaviness across CRAN and Bioconductor.

Main Results:

• Identified top R packages with significant 'dependency heaviness' contributions.
• The 'pkgndep' package offers an intuitive approach to dependency analysis.
• Analysis revealed patterns of dependency complexity in major R repositories.

Conclusions:

• Dependency heaviness provides a valuable measure for understanding and managing R package dependencies.
• Optimizing the use of 'heavy' parent packages can reduce overall complexity.
• The findings aid in developing more manageable and robust bioinformatics software ecosystems.